In commercial plant breeding the production of hybrid seed is very important. Plants grown from hybrid seed are generally very uniform, and they benefit from heterosis (hybrid vigour), which can lead to a significant increase in yield and/or performance when compared to the parental lines of the hybrid, or to outcrossing (open-pollinated) lines. Typically the parental lines used for hybrid seed production are inbred, which implies that their genomes are largely homozygous. The combination of two largely homozygous genomes into a hybrid leads to a high degree of heterozygosity, if both parental lines were genetically unrelated or not closely related.
Efficient hybrid seed production in plant species that are able to self-fertilise requires adequate measures to prevent self-fertilisation of the plants on which hybrid seeds are to be produced. Various strategies have been developed to achieve this, and to obtain an efficient hybrid seed production setup. However, the complexity and amount of labour required for each of these strategies varies greatly.
A strategy that naturally occurs in certain plant species is the physical separation of male and female reproductive organs in separate flowers, either on separate plants (dioecious species) or on the same plant (monoecious species). This system naturally promotes outcrossing, and it can be easily taken advantage of for hybrid seed production.
Another natural strategy is self-incompatibility, which has e.g. been extensively studied in Brassica species. In this case, pollen is physically unable to fertilise egg cells from the same plant. The precise mechanism of the incompatibility interaction can differ. Either pollen hydration or germination is prevented, or pollen tube growth through the style is inhibited by the female tissues, or the pollen tube is not attracted to ripe ovules, or the sperm nuclei are unable to merge with the egg cell nucleus to form a viable zygote. Again, this naturally occurring system is very efficient and useful for preventing self-fertilisation, and for promoting outcrossing.
Another method for preventing selfing, which is typically used in e.g. maize, is the mechanical elimination of all male flowers (detasseling). The only flowers remaining on the plant are female, and these can be manually pollinated with pollen from a selected paternal line, in order to obtain ears with exclusively hybrid kernels.
In plant species with hermaphroditic flowers (producing both ovules and pollen grains within the same flower), a common strategy for preventing selfing is emasculation by mechanical removal of anthers and/or pollen prior to anthesis. When the anthers are mechanically removed before the pollen grains are released from the loculi and/or before the filament has extended far enough to match the height of the stigma, selfing is efficiently prevented. Subsequently the female reproductive parts of the emasculated flower are allowed to mature normally, after which pollen grains from a selected father plant can be deposited on the stigma, in order to obtain exclusively hybrid seeds from the cross. Especially for commercial-scale applications this method is however very labour-intensive and not 100% reliable: if anthers are removed in a slightly too late developmental stadium or if one anther is accidentally not removed, this can lead to a mixed seed set, consisting of hybrid and maternal seeds. This results in non-uniformity of the commercial seed batch, which is undesired for customers who expect uniform and consistently superior seeds, and it brings inbred mother lines of hybrid varieties into commerce, which is undesired for the breeding company. A 100% reliable hybrid system is therefore desirable.
Another approach is to induce male sterility by means of chemicals. This so-called male gametocide can be achieved by treatment with e.g. gibberellins (in rice and maize), sodium methyl arsenate (in rice), or maleic acid (in wheat and onion). Disadvantages of this approach are the fact that this male sterility is not inheritable as it does not result from a genetic determinant present in the plant's genome, and that chemical treatment is labour intensive and not 100% reliable.
Another category of mechanisms through which selfing can be prevented, is termed genetic male sterility. Here three different approaches can be distinguished: genetic-engineered male sterility (transgenic MS), cytoplasmic male sterility (CMS) and genic male sterility (GMS). Transgenic MS comprises all approaches that use a transgene to ensure that pollen grains are unable to fertilise ovules, and that either lead to the death of pollen grains prior to anthesis, or to the dysfunctionality of pollen grains at anthesis. A well-known example is the reversible Barnase/Barstar system, wherein the Barnase enzyme is transgenically expressed in the tapetum, which leads to pollen sterility. However, when the Barstar protein is co-expressed in the tapetum, it blocks Barnase activity and restores pollen fertility (Mariani et al., 1990, Nature 347: 737-41).
Cytoplasmic male sterility (CMS) is a type of sterility that is under control of extra-nuclear, cytoplasmic factors, more precisely of plastid origin. Usually mutations in the mitochondrial genome underlie CMS, and they typically inherit in a maternal fashion. Routine hybrid seed production with CMS lines requires the use of maintainer and restorer lines, which complicates the process and increases the costs and time required for commercial hybrid seed production.
Genic male sterility (GMS) encompasses a nuclear influence on male fertility, in contrast to cytoplasmic influences which are caused by organellar factors. Due to e.g. a mutation in a nuclear gene the plant does not produce viable and/or functional pollen grains or male spores, and/or it is unable to disperse its pollen due to e.g. non-dehiscence of its anthers.
The Asteraceae family—also known as the Compositae family—is one of the largest extant plant families. It comprises various commercially important crop plants, such as sunflower (Helianthus), lettuce (Lactuca), endive, witloof and radicchio (Cichorium), artichoke (Cynara), and many ornamental plants such as Chrysanthemum, Tagetes, Gerbera and Zinnia. 
In the Compositae family only few types of male sterility have been developed for commercial exploitation and hybrid seed production. Emasculation is very difficult in this family, due to the composite nature of the inflorescences. In e.g. endive and witloof, self-pollinaton is normally prevented by spraying an inflorescence with water, to flush away the pollen. Timing is crucial, because this spraying needs to be done immediately after opening of the anthers, before the stigma splits into two curving parts that protrude beyond the anthers. After spraying the inflorescence has to be blown dry and has to be allowed to develop further, before pollen grains from a selected father can be deposited onto the stigma. It is critical to choose the optimal moment for spraying with water: if pollen are removed too late some self-pollination will already have occurred, and if pollen are removed before the anthers are fully opened, pollen grains will remain present and may cause self-pollination at a later stage. Every day there is only a limited time window during which this spraying can be done, and its timing depends strongly on light and temperature conditions. The development of an efficient, reversible male sterility system in the Compositae family would thus greatly facilitate breeding in crop species belonging to this family.
Hybrid sunflowers are available and can be produced using various methods described above. However for e.g. endive a hybrid system is not yet available at all, and a major problem is the very limited genetic variation within this cultivated crop. If an efficient hybrid system would be available for endive, this could be used to increase the genetic variation, as well as provide additional benefits through heterotic effects. Interspecific crosses with e.g. witloof (Cichorium intybus) are possible, and these could be used to introduce foreign genetic material into endive germplasm, but this would lead to complications at the genetic level, namely the fact that the resulting progeny would no longer be purely endive, but a mixture of endive and witloof.
In witloof a recessive GMS trait has previously been created through a transposon insertion in a homologue of the DYT1 gene of Arabidopsis (Quillet et al., 2011, Cloning and characterization of nuclear male sterility 1 (nms1) in chicory. However, in practice it is very difficult to transfer this transposon-based trait to endive (Cichorium endivia). Also, although the transposon may spontaneously be excised from the DYT1 gene homologue and thus potentially restore fertility, a researcher is not able to easily, predictably and consistently reverse this male sterility trait in witloof whenever he wishes to do so.
CMS has also been created in witloof, by combining the nuclear genome from C. intybus with cytoplasm from sunflower (Hehanthus annuus). However, again this is not a reversible male sterility, and the perpetuation of the trait requires maintainer and restorer lines.
In lettuce, male sterility has been described in the prior art, and pollination for obtaining hybrid seeds can e.g. be achieved with bees (U.S. Pat. No. 7,569,743). Dominant GMS is available (resulting from the MS7 mutation), as well as CMS (through combination of the nuclear genome of Lactuca sativa with the cytoplasm of sunflower). However, none of these traits is reversible, and it thus requires more efforts to maintain the male sterile mother lines in any of those cases.
One could imagine a transgenic approach to obtain reversible male sterility in Compositae plants, e.g. with the Barnase/Barstar system, but such transgenic plants would have a “regulated” status, which is undesired in e.g. the European market, and this would necessitate large extra deregulation expenses to obtain market approval. The cost for bringing hybrid seeds resulting from the use of such a transgenic reversible male-sterile Compositae plant to the market would thus become quite high.
Citation or identification of any document in this application is not an admission that such document is available as prior art to the present invention.